基于S型曲线的模糊故障树及其应用

Fuzzy Fault Tree and Its Application Based on S Curve

在传统模糊故障树模型综合评估中很少考虑专家可信度的影响,处理专家意见时很少考虑专家评估的可信度。文中提出基于置信度的S型曲线模糊数确定方法,引入专家权威性系数和一致性系数确定专家的置信度。运用概率偏差方法求解S型曲线的可调参数,基于可调参数和置信度来校准故障树基本事件的模糊失效概率,增加了基本事件模糊数确定的灵活性。以作战系统为例进行了实验,结果表明,针对同一基本事件各专家评估分歧较大,在各基本事件之间评估分布均匀的情况下,在未经置信处理之前各基本事件得到的最终失效概率波动比较明显,而经置信处理后的各失效概率趋于稳定,验证了改进的模糊故障树在模糊失效概率确定中结果更符合客观实际,增大了模糊故障树诊断的准确性。

Aiming at the problem of the comprehensive evaluation of the traditional fuzzy fault tree model,the reliability of the expert e- valuation is seldom considered in the process of dealing with the expert opinion. Fuzzy number determination method based on the confidence degree of S type curve is proposed,introduction of the expert＇ s authority coefficient and the consistency coefficient. The probability deviation method is used to get the adjustable parameters of the S type curve, and the fuzzy failure probability of the basic events for the fault tree can be calibrated based on the adjustable parameters and the confidence level. Experiments are carried out in the combat system, whose results show that the assessment of the same basic events is quite different when the assessment of the distribution of the basic events is even, the final failure probability of each basic event is more obvious, and the probability of failure probability is more stable and verified that improved fuzzy fault tree is more consistent with the objective reality, and the accuracy of the fuzzy fault tree diagnosis is in- creased.